Direct observation of long-lived radical pair between flavin and guanine in single- and double-stranded DNA-oligomers

Abstract

The mechanism by which cryptochrome (CRY) proteins are capable of sensing weak magnetic fields (e.g., the geomagnetic field: ~50 μT) was suggested to be mediated by spin-correlated radical pairs (SCRPs) comprising a flavin adenine dinucleotide (FAD) radical and a tryptophan (Trp) radical which are formed simultaneously by light-induced electron transfer (ET). Here, we provide evidence for direct photoinduced ET that leads to long-lived SCRPs comprising a flavin (Fl) radical and a guanine (G) radical in flavin-tethered single- and double-stranded DNA oligomers by using time-resolved electron paramagnetic resonance (TREPR) spectroscopy. Transient absorption (TA) spectroscopy and its magnetic field effect (MFE) identified RP generation via a triplet-state precursor, in contrast to RP generation via a singlet-state precursor in CRY. Our findings of RPs in Fl-DNA oligomers having microsecond-long lifetimes and capable of exerting a large MFE at room temperature may significantly impact on our understanding of biological magnetoreception.

Scheme 1. Photoinduced ET processes of guanine (G) and flavin (Fl).

After the processes a) and b), sequential electron/hole transfer processes between G bases are possible in G polyads. The processes a) and b) are inhibited by the presence of an indole moiety. Assuming that intrastrand reactions are almost exclusive, process e) should be negligible, see the main text^§.

Fig. 1. Quenching efficiencies of flavin fluorescence in the ODNs.

Fluorescence intensities of flavin in the ODNs (50 μM, I) relative to the parent flavin molecule (50 μM, I₀) measured at 16 °C. Excitation: 450 nm and emission: 520 nm. The averages and the standard deviations (in parentheses) of the respective measurements of three times are plotted.

Fig. 2. TREPR spectra of ss-Fl–DNA(G4) and G-to-I substituent as a negative control.

50 μM solutions of (a) ss-Fl–DNA(G4) and (b) ss-Fl–DNA(I1G3), recorded 1.5 μs after pulsed laser excitation (at 450 nm) at 5 °C. Experimental spectra are shown as drawn lines, spectral simulations as dashed lines. The parameters entering the simulations can be found in Supplementary Table S2.

Fig. 3. TREPR spectra of ss-Fl–DNA(G1I3) and the temperature dependence.

50 μM solution of ss-Fl–DNA(G1I3), recorded 1.5 μs after pulsed laser excitation (at 450 nm), (a) in the solution state at 5 °C and (b) in the frozen solution state at −123 °C. Experimental spectra are shown as drawn lines, spectral simulations as dashed lines. The parameters entering the simulations can be found in Supplementary Table S2.

Fig. 4. TREPR spectra of ds-Fl–DNA(G4)/N–DNA and the suppressing effect.

50 μM solutions of (a) ds-Fl–DNA(G4)/N–DNA, (b) ds-Fl–DNA(I1G3)/N–DNA and (c) ds-Fl–DNA(G4)/Ind–DNA, recorded 1.5 μs after pulsed laser excitation (at 450 nm) at 5 °C. Experimental spectra are shown as drawn lines, spectral simulations as dashed lines. The parameters entering the simulations can be found in Supplementary Table S2.

Fig. 5. Time profiles of transient absorption (TA) of ss-Fl–ODN and the magnetic field effect (MFE).

a Time profiles of transient absorption (TA) for a 50 μM solution of ss-Fl–DNA(G4) and (b) Magnetic field effect on the TA signal averaged in the range from 0.2 to 0.5 μs after laser pulses. The TA signals are recorded monitoring the 532 nm diode laser after a pulse excitation by OPO (450 nm) at room temperature ( ~23 °C). The magnetic field is scanned from -28 to +28 mT and repeated 10 times. The error bar is calculated by the standard error over 10 times data. No error bar is on the zero-field data because the data are calculated relative to the data at zero field at each scan. The filled circles are for ss-Fl–DNA(G4) and open circles are for ss-Fl–DNA(I4) as a control experiment.